Questions and Answers on System Suitability Tests (SST) - Part 2

During the ECA Live Online Training “System Suitability Tests (SST) and Troubleshooting for HPLC Methods” on 04 November 2025, a number of interesting and practice-oriented questions were raised by the participants. These questions were answered in writing by the speakers’ team after the event. To share some of the key insights with a broader audience, we have compiled a selection of these Q&As.

Please find below Part 2 of the questions with answers provided by Dr Gerd Jilge. Part 1 was published in December, and Part 3 will be published in a few months.

Part 2

2.1 – For long sequence run times, you mentioned “repeat the SST”. Would it be acceptable, for example, to perform five SST injections at the beginning of the sequence and only one standard injection at the end to calculate the %RSD?
Regarding the proposal to perform the SST by injecting 5x at the beginning and 1x at the end I have the following concerns:
The analysis cannot be started before the SST is fulfilled. That means that the RSD for 5 injections has to be calculated. Therefore, the 6th SST injection cannot be included into the calculation of the other 5 injections because the analysis will be started before.
Otherwise, a single SST injection does not make sense. No RSD calculation is possible.
In such cases using long runs I recommend to perform the SST at the beginning and in the middle of the sequence (if necessary).
The best solution, however, would be to perform the SST before starting the analysis and to control the analytical sequence by injecting additional control standards. Control standards can be injected in a regular manner during the sequence.

2.2 – If the SST fails, for example, twice and a third attempt also fails, at what point should testing be stopped and a formal investigation initiated? Is this situation addressed in any guideline or regulatory document? Should such an investigation be handled in a manner similar to an OOS investigation (Phase I)?
An investigation is necessary after each failed SST and should be started after the 1st failed SST! The FDA OOS guidance describes it as follows: "If reference standard responses indicate that the system is not functioning properly, all of the data collected during the suspect time period should be properly identified and should not be used. The cause of the malfunction should be identified and, if possible, corrected before a decision is made whether to use any data prior to the suspect period."
It is not an investigation like an investigation after getting an OOS result (but please be careful because it will change if a part or the complete of the analysis has been performed with a SST failure).

2.3 – If the analysis is long, how often should the SST be repeated? Also, do the standard injections need to be repeated?
A SST can be repeated after very long chromatographic runs (e.g. 24 hours). However, it might be much better to perform control standard after a certain number of injections. Please be careful because in case of a failed SST at the end of a chromatographic analysis, the analytical results cannot be used (see also question 2.2).

2.4 – How should the precision or similarity between duplicate sample preparations (single injection each) be evaluated for assay and related substances analyses?
In case of the validation experiments the best way to evaluate the sampling strategy is to perform the intermediate precision varying analysts, days, equipment etc. Furthermore, for each sample preparation at least 3 injections are recommended in order to define the variability of the injection repeatability. Based on these data it is possible to define a suitable number of sample preparations and the number of injections (also relevant for double sample preparations with one injection each).
The strategy can be used for assay as well for the determination of impurities (e.g. by spiking with reference substances).

2.5 – In our laboratory, we use a so-called “first injection” to condition the column. This first injection is typically a buffer or solvent and is performed for conditioning purposes only; the data are not evaluated. However, we have no clear evidence that this practice is effective. What is your opinion on this approach?
In general, the first injection is performed using a blank solution to exclude effects which might be relevant e.g. in case of the determination of impurities. These data are not used for the calculation of the SST but may indicate impurities from the system (e.g. mobile phase). In such cases it might be useful to set peak areas as a limit for blind peaks in order to exclude an impact on the integration of the analytes during analysis.